This paper presents a novel hybrid method for the determination of intrinsic geometric parameters in the case of a fundus camera from a set of acquired images. The main idea is to integrate two camera calibration methods, the most popular in the field of computer vision, Agapito and Zhang, by means of the optimization procedure. In comparison with traditional techniques suffering from a possible degeneration of algorithms, the proposed approach gives satisfactory results and has demonstrated the expected performance improvements in terms of efficiency and robustness.
Volume 12 | 05-Special Issue
Pages: 166-172
DOI: 10.5373/JARDCS/V12SP5/20201745